INDEX

Explanations

the word "won't" with high activation values

oai_token-act-pair · gpt-3.5-turbo

negations or words indicating refusal or denial

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

 behavi

-0.77

CVE

-0.72

 Reloaded

-0.72

Reviewer

-0.72

 examiner

-0.70

 Palestin

-0.70

HTTP

-0.67

 Moroc

-0.67

Hardware

-0.66

 ventilation

-0.65

POSITIVE LOGITS

weet

1.01

ardless

0.91

acular

0.91

ruck

0.91

urtle

0.91

ravis

0.86

aylor

0.81

otally

0.80

itles

0.79

rees

0.79

Activations Density 0.032%

the word "won't" with high activation values

negations or words indicating refusal or denial

No Comments

No Known Activations