INDEX

Explanations

adjective phrases describing extremes or intensity

oai_token-act-pair · gpt-3.5-turbo

statements that describe resulting conditions or 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

aiden

-0.63

olly

-0.61

AST

-0.60

 Shiva

-0.60

yne

-0.59

 Calm

-0.59

å°Ĩ

-0.58

 Shiite

-0.58

âĸĪâĸĪâĸĪâĸĪâĸĪâĸĪâĸĪâĸĪ

-0.58

 Witch

-0.57

POSITIVE LOGITS

borgh

0.79

rouse

0.73

0.72

ptions

0.72

itated

0.71

enf

0.64

akespe

0.62

ymes

0.61

bed

0.61

smanship

0.61

Activations Density 0.302%

adjective phrases describing extremes or intensity

statements that describe resulting conditions or conclusions

No Comments

No Known Activations