INDEX

Explanations

words related to trying out new methods or ideas

oai_token-act-pair · gpt-3.5-turbo

instances of the word "experiment" and its variations, indicating a focus on experimentation across various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.3.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.3.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

othy

-0.68

translation

-0.67

cut

-0.66

si

-0.66

sama

-0.64

HCR

-0.64

say

-0.64

byn

-0.63

vation

-0.62

trans

-0.62

POSITIVE LOGITS

 experimenting

1.11

 experimented

0.98

 withd

0.94

 experimentation

0.89

iments

0.85

 tink

0.84

imental

0.82

ienced

0.78

redients

0.77

isher

0.74

Activations Density 0.015%

words related to trying out new methods or ideas

instances of the word "experiment" and its variations, indicating a focus on experimentation across various contexts

No Comments

No Known Activations