INDEX

Explanations

phrases indicating evaluation or judgement

oai_token-act-pair · gpt-3.5-turbo

expressions of subjective experience or personal feelings

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

neutral

-0.79

bright

-0.64

planes

-0.64

folios

-0.61

raltar

-0.61

usha

-0.60

icent

-0.59

allery

-0.58

ãĥĥãĥī

-0.58

winner

-0.57

POSITIVE LOGITS

 compared

0.68

arella

0.66

mouth

0.64

inventoryQuantity

0.64

 sounding

0.63

Feel

0.62

 actually

0.61

Laughs

0.59

 indeed

0.59

externalActionCode

0.58

Activations Density 0.134%

phrases indicating evaluation or judgement

expressions of subjective experience or personal feelings

No Comments

No Known Activations