INDEX

Explanations

phrases indicating possession or ownership

oai_token-act-pair · gpt-3.5-turbo

phrases emphasizing possession or individuality

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

antle

-0.76

Trade

-0.72

reading

-0.72

anwhile

-0.70

itement

-0.70

mer

-0.67

obe

-0.66

MER

-0.66

 Bullets

-0.65

olic

-0.64

POSITIVE LOGITS

 unique

0.87

 version

0.85

 distinct

0.81

 custom

0.79

 distinctive

0.78

 separate

0.74

 turf

0.74

 proprietary

0.73

 branded

0.72

 idiosyncr

0.71

Activations Density 0.031%

phrases indicating possession or ownership

phrases emphasizing possession or individuality

No Comments

No Known Activations