INDEX

Explanations

words related to buildings, structures, or physical locations

oai_token-act-pair · gpt-3.5-turbo

words related to the concept of being "ed" or "conditioned," often referring to modifications or states

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

 Kiss

-0.67

Story

-0.64

 derog

-0.63

 Spears

-0.63

 Royale

-0.63

Aux

-0.63

Nun

-0.63

 Union

-0.62

 Anthem

-0.62

 sight

-0.61

POSITIVE LOGITS

ict

1.14

ifice

1.13

gew

1.13

uce

1.08

icts

1.01

icy

0.99

dy

0.99

ging

0.98

nesday

0.97

monton

0.96

Activations Density 0.013%

words related to buildings, structures, or physical locations

words related to the concept of being "ed" or "conditioned," often referring to modifications or states

No Comments

No Known Activations