INDEX

Explanations

statements related to the effectiveness or ineffectiveness of certain actions or policies

oai_token-act-pair · gpt-3.5-turbo

arguments related to the effectiveness of various policies or initiatives

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.10.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Printed

-0.85

 brood

-0.73

 handwritten

-0.73

 guessing

-0.68

 Naked

-0.67

 ridicule

-0.67

 disbelief

-0.66

 Fancy

-0.66

Fn

-0.65

 Jere

-0.64

POSITIVE LOGITS

acea

1.31

 beneficial

1.21

 effic

1.14

 allev

1.12

 reducing

1.10

 sustainable

1.10

 deterrent

1.09

 alleviate

1.06

 deterrence

1.05

 boosting

1.04

Activations Density 0.603%

statements related to the effectiveness or ineffectiveness of certain actions or policies

arguments related to the effectiveness of various policies or initiatives

No Comments

No Known Activations