INDEX

Explanations

prepositions indicating a point or direction in physical space

oai_token-act-pair · gpt-3.5-turbo

phrases that denote progression or continuity in a narrative context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

assic

-0.73

 spoil

-0.68

amen

-0.62

rouse

-0.59

 tresp

-0.59

 unres

-0.59

 rapt

-0.58

 offensive

-0.57

 mess

-0.57

ilty

-0.57

POSITIVE LOGITS

neath

0.94

ward

0.88

itialized

0.81

 Racer

0.79

lash

0.78

assis

0.76

ctr

0.73

math

0.72

coming

0.72

wards

0.72

Activations Density 0.109%

prepositions indicating a point or direction in physical space

phrases that denote progression or continuity in a narrative context

No Comments

No Known Activations