INDEX

Explanations

phrases related to situations or events being resolved or working out positively

oai_token-act-pair · gpt-3.5-turbo

phrases indicating the overall state of situations or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 9-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.9.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.9.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

ktop

-0.71

¿½

-0.68

eph

-0.66

edia

-0.64

orst

-0.64

vernment

-0.63

many

-0.61

rique

-0.60

 outnumbered

-0.60

Reader

-0.59

POSITIVE LOGITS

 except

1.08

 smoothly

0.83

except

0.83

 revolves

0.81

perfect

0.73

 melts

0.72

together

0.72

 neatly

0.67

 together

0.67

 imaginable

0.67

Activations Density 0.266%

phrases related to situations or events being resolved or working out positively

phrases indicating the overall state of situations or events

No Comments

No Known Activations