INDEX

Explanations

words or phrases related to additional information or related topics, possibly in a list format

oai_token-act-pair · gpt-3.5-turbo

the phrase "Also" followed by significant information or references

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

ience

-0.67

itiveness

-0.66

atics

-0.65

acious

-0.64

ILLE

-0.64

ylum

-0.64

slot

-0.63

itness

-0.61

Â£

-0.61

ptoms

-0.61

POSITIVE LOGITS

 noteworthy

0.91

 notable

0.79

 Known

0.79

 Called

0.75

epad

0.71

 includes

0.70

 optionally

0.68

 included

0.66

 notor

0.65

 includ

0.65

Activations Density 0.057%

words or phrases related to additional information or related topics, possibly in a list format

the phrase "Also" followed by significant information or references

No Comments

No Known Activations