INDEX

Explanations

instances where something is assumed or appears to be a certain way

oai_token-act-pair · gpt-3.5-turbo

phrases indicating perceptions or conclusions drawn by individuals

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

-0.62

complete

-0.57

onica

-0.56

pat

-0.56

 Submit

-0.55

PM

-0.55

-0.54

HM

-0.53

 collect

-0.53

idem

-0.52

POSITIVE LOGITS

 seemed

3.16

 seems

2.09

 seem

2.06

 sounded

2.02

 appeared

2.02

 resembled

1.86

 mattered

1.66

 tended

1.56

 looked

1.55

 lasted

1.45

Activations Density 0.013%

instances where something is assumed or appears to be a certain way

phrases indicating perceptions or conclusions drawn by individuals

No Comments

No Known Activations