INDEX

Explanations

proper nouns, specifically names of places and people

oai_token-act-pair · gpt-3.5-turbo

words that are frequently occurring in a context related to Brussels

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ĸļ

-0.81

©¶æ

-0.71

 Uriel

-0.61

 Pradesh

-0.61

 Azerbai

-0.60

 PROG

-0.59

 nitrogen

-0.59

 newsp

-0.58

 lodging

-0.57

 thor

-0.57

POSITIVE LOGITS

sels

0.99

hemy

0.98

itte

0.94

veland

0.94

gger

0.93

olini

0.92

zinski

0.89

swick

0.89

Ã§a

0.85

illin

0.82

Activations Density 0.095%

proper nouns, specifically names of places and people

words that are frequently occurring in a context related to Brussels

No Comments

No Known Activations