INDEX

Explanations

words related to significant choices or actions

oai_token-act-pair · gpt-3.5-turbo

references to significant decisions

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

vae

-0.77

english

-0.72

 havoc

-0.67

amen

-0.67

icas

-0.67

 tremend

-0.67

outh

-0.67

uum

-0.66

ighth

-0.65

ingers

-0.65

POSITIVE LOGITS

 makers

1.01

jar

0.94

 maker

0.86

 decision

0.83

making

0.81

 ACTIONS

0.80

maker

0.78

 decisions

0.77

makers

0.75

lessness

0.71

Activations Density 0.037%

words related to significant choices or actions

references to significant decisions

No Comments

No Known Activations