INDEX

Explanations

words related to measurements or numerical values

oai_token-act-pair · gpt-3.5-turbo

references to trips or rounds, particularly in relation to travel or transportation 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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

PsyNetMessage

-0.78

aline

-0.77

sworth

-0.69

maid

-0.69

iments

-0.68

oir

-0.67

ibo

-0.65

eele

-0.64

ieri

-0.63

iment

-0.63

POSITIVE LOGITS

Robin

0.84

trip

0.73

©¶æ

0.68

Bottom

0.66

Rect

0.65

estate

0.65

 kernels

0.65

lege

0.65

CBC

0.64

Circ

0.64

Activations Density 0.070%

words related to measurements or numerical values

references to trips or rounds, particularly in relation to travel or transportation contexts

No Comments

No Known Activations