INDEX

Explanations

actions or responses to situations

oai_token-act-pair · gpt-3.5-turbo

phrases indicating responses or actions taken by entities

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

 contiguous

-0.69

 ILCS

-0.69

 ancestor

-0.66

 slaves

-0.64

 Located

-0.64

igs

-0.62

ulhu

-0.62

sterdam

-0.61

ãĥ´

-0.60

 Signed

-0.59

POSITIVE LOGITS

0.78

 disapproval

0.75

 barrage

0.73

itely

0.70

cold

0.67

 onslaught

0.67

 prov

0.66

 voic

0.66

 chorus

0.65

 dismissal

0.65

Activations Density 0.215%

actions or responses to situations

phrases indicating responses or actions taken by entities

No Comments

No Known Activations