INDEX

Explanations

terms related to "models" as in examples, representations, or role models

oai_token-act-pair · gpt-3.5-turbo

references to models or prototypes in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

 Citation

-0.74

eways

-0.69

 crest

-0.66

poral

-0.65

ifact

-0.61

tions

-0.61

oS

-0.60

 reserved

-0.60

 strap

-0.60

 recess

-0.59

POSITIVE LOGITS

aos

0.71

rha

0.69

hur

0.67

berus

0.66

uci

0.65

 Fenrir

0.65

ousing

0.65

Mary

0.65

getic

0.65

Downloadha

0.65

Activations Density 0.000%

terms related to "models" as in examples, representations, or role models

references to models or prototypes in various contexts

No Comments

No Known Activations