INDEX

Explanations

verbs related to deduction, such as "infer" and "extrapolate."

oai_token-act-pair · gpt-3.5-turbo

terms related to reasoning and drawing conclusions

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

 deeds

-0.78

zilla

-0.70

sche

-0.67

Benz

-0.65

 deed

-0.64

mk

-0.64

nil

-0.62

locking

-0.59

 Directive

-0.59

lock

-0.59

POSITIVE LOGITS

 infer

0.87

iour

0.87

 extrap

0.85

 inferred

0.85

iors

0.81

 glean

0.79

itial

0.79

olate

0.78

æĥ

0.78

»

0.77

Activations Density 0.045%

verbs related to deduction, such as "infer" and "extrapolate."

terms related to reasoning and drawing conclusions

No Comments

No Known Activations