INDEX

Explanations

phrases related to being "right inside" something and giving a sense of scale

oai_token-act-pair · gpt-3.5-turbo

references to "D" and "ere" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

OWS

-0.83

ows

-0.78

orously

-0.77

eston

-0.74

owed

-0.73

å°Ĩ

-0.73

gag

-0.69

ORE

-0.69

edit

-0.67

hips

-0.66

POSITIVE LOGITS

ndum

1.21

nda

1.00

tta

0.96

lict

0.94

mma

0.92

cki

0.90

gger

0.90

nder

0.90

cht

0.88

lda

0.87

Activations Density 0.028%

phrases related to being "right inside" something and giving a sense of scale

references to "D" and "ere" in various contexts

No Comments

No Known Activations

phrases related to being "right inside" something and giving a sense of scale

references to "D" and "ere" in various contexts

No Comments

No Known Activations