INDEX

Explanations

positive feedback or praise

oai_token-act-pair · gpt-3.5-turbo

overall ratings or assessments of experiences or items

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

ffe

-0.72

 grandchildren

-0.69

 newsletters

-0.65

lil

-0.65

 toile

-0.64

href

-0.64

 attendant

-0.63

aciously

-0.63

lÃ©

-0.62

adr

-0.62

POSITIVE LOGITS

 Thoughts

0.85

 Length

0.77

 Ratio

0.73

apse

0.72

Overall

0.71

 Summary

0.70

 rating

0.70

hesis

0.69

 Rating

0.69

 Overall

0.68

Activations Density 0.032%

positive feedback or praise

overall ratings or assessments of experiences or items

No Comments

No Known Activations