INDEX

Explanations

phrases related to causation and responsibility

oai_token-act-pair · gpt-3.5-turbo

phrases that indicate causation or responsibility

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

atography

-0.74

atri

-0.72

arel

-0.71

sha

-0.69

efer

-0.69

arious

-0.68

 Technique

-0.66

heit

-0.65

ography

-0.65

ynchronous

-0.65

POSITIVE LOGITS

 instability

0.96

 deaths

0.91

 exacerb

0.90

 worsened

0.88

 worsening

0.87

 relapse

0.86

why

0.86

 susceptibility

0.85

 outbreaks

0.85

 discrepancies

0.84

Activations Density 0.274%

phrases related to causation and responsibility

phrases that indicate causation or responsibility

No Comments

No Known Activations