INDEX

Explanations

file formatting or programming language-related terms

oai_token-act-pair · gpt-3.5-turbo

keys or indicators related to error messages and programming elements

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

çĶ°

-0.66

ONSORED

-0.66

ktop

-0.62

Ô

-0.61

\\\\\\\\

-0.61

è»

-0.60

iHUD

-0.58

hower

-0.56

ISION

-0.55

Ö

-0.55

POSITIVE LOGITS

 respective

0.65

 backgrounds

0.61

blogs

0.55

 ranging

0.55

 including

0.55

 apiece

0.55

including

0.54

 folders

0.53

 individually

0.52

 descriptions

0.51

Activations Density 2.245%

file formatting or programming language-related terms

keys or indicators related to error messages and programming elements

No Comments

No Known Activations