INDEX

Explanations

links or references in a text

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the location of information or resources

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

 evacuation

-0.76

 judgement

-0.73

 vaccinations

-0.73

 judgment

-0.72

 separation

-0.70

 abandonment

-0.69

 diagnosis

-0.67

 clipboard

-0.67

 correctness

-0.67

 precaution

-0.67

POSITIVE LOGITS

ById

0.86

lopp

0.84

swick

0.82

 HERE

0.81

resso

0.81

 Tradable

0.81

(),

0.79

EGA

0.70

GG

0.70

 here

0.70

Activations Density 0.043%

links or references in a text

phrases indicating the location of information or resources

No Comments

No Known Activations