INDEX

Explanations

phrases related to progress or advancement

oai_token-act-pair · gpt-3.5-turbo

phrases related to progress or improvement over time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

until

-0.62

cock

-0.58

 Remain

-0.58

ashington

-0.58

 McGill

-0.57

 restraint

-0.56

 risked

-0.54

 Kard

-0.54

 begin

-0.54

DF

-0.53

POSITIVE LOGITS

 unexpectedly

0.94

 handy

0.81

 fruition

0.78

 leaps

0.77

 naturally

0.76

OX

0.72

 sidx

0.72

soon

0.71

 blows

0.70

 spontaneously

0.69

Activations Density 0.166%

phrases related to progress or advancement

phrases related to progress or improvement over time

No Comments

No Known Activations