INDEX

Explanations

mentions of different seas or bodies of water

oai_token-act-pair · gpt-3.5-turbo

references to "seas" 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 @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

gee

-0.79

ieu

-0.76

ocular

-0.76

isma

-0.70

actively

-0.68

achment

-0.67

igun

-0.67

okia

-0.66

omination

-0.66

ison

-0.66

POSITIVE LOGITS

hell

1.04

 Seas

0.94

cream

0.85

 seas

0.84

 Islands

0.80

manship

0.78

lass

0.77

mares

0.76

ource

0.76

orthy

0.75

Activations Density 0.034%

mentions of different seas or bodies of water

references to "seas" in various contexts

No Comments

No Known Activations