INDEX

Explanations

mentions of technical instructions or guidelines

oai_token-act-pair · gpt-3.5-turbo

occurrences of notable or significant events or references

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

''.

-0.94

)</

-0.87

Failure

-0.80

ocide

-0.79

 regrets

-0.79

.''

-0.78

\<

-0.76

îĢ

-0.74

error

-0.73

exit

-0.73

POSITIVE LOGITS

 dudes

0.82

 pairing

0.79

 themed

0.77

 showcasing

0.72

 vibe

0.72

 excited

0.72

 showcased

0.72

 paired

0.72

 gorgeous

0.69

 flashy

0.68

Activations Density 1.043%

mentions of technical instructions or guidelines

occurrences of notable or significant events or references

No Comments

No Known Activations