INDEX

Explanations

instances where a piece of information or observation is being mentioned or relayed

oai_token-act-pair · gpt-3.5-turbo

instances of the word "noted"

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

quer

-0.93

onz

-0.72

icum

-0.72

ravel

-0.69

atom

-0.68

adesh

-0.67

shell

-0.66

ibur

-0.65

imeters

-0.64

wra

-0.63

POSITIVE LOGITS

 inconsistencies

0.97

 similarities

0.92

 discrepancies

0.87

 approving

0.87

how

0.87

 noting

0.86

 favorably

0.84

 prominently

0.82

 faults

0.79

 inaccur

0.77

Activations Density 0.067%

instances where a piece of information or observation is being mentioned or relayed

instances of the word "noted"

No Comments

No Known Activations