INDEX

Explanations

phrases indicating setting something aside or putting aside concerns

oai_token-act-pair · gpt-3.5-turbo

phrases encouraging the setting aside of differences or distractions

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

ENTS

-0.63

wei

-0.61

jah

-0.61

enos

-0.60

 enthusi

-0.60

amen

-0.59

 promot

-0.59

enez

-0.58

 pestic

-0.57

arming

-0.57

POSITIVE LOGITS

 altogether

0.77

 aside

0.77

 entirely

0.75

 Unch

0.71

heid

0.69

 differences

0.66

thal

0.65

for

0.65

 until

0.64

urities

0.62

Activations Density 0.031%

phrases indicating setting something aside or putting aside concerns

phrases encouraging the setting aside of differences or distractions

No Comments

No Known Activations