INDEX

Explanations

references to the word "Lemon" and its variations

oai_token-act-pair · gpt-3.5-turbo

specific names or references, particularly those associated with a particular individual or character

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

ãĥ¼ãĥĨ

-0.84

ãģ¦

-0.72

================================================================

-0.65

 à¨

-0.64

CLASSIFIED

-0.64

ÙĲ

-0.63

Christ

-0.63

ãĥ¼ãĥĨãĤ£

-0.62

 Sakura

-0.62

 Tant

-0.62

POSITIVE LOGITS

ike

1.01

ongo

0.96

enko

0.95

rade

0.94

isine

0.94

iris

0.91

roots

0.90

eny

0.89

uz

0.87

rov

0.87

Activations Density 0.027%

references to the word "Lemon" and its variations

specific names or references, particularly those associated with a particular individual or character

No Comments

No Known Activations