INDEX

Explanations

expressions related to making judgments or judgments themselves

oai_token-act-pair · gpt-3.5-turbo

occurrences of the word "judgment" and its variants

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

href

-0.74

KC

-0.74

ieri

-0.71

mers

-0.70

oples

-0.69

hedral

-0.69

heny

-0.68

tail

-0.67

oppy

-0.66

chell

-0.66

POSITIVE LOGITS

 judgement

1.25

 judgment

1.24

 judgments

1.05

 Judgment

1.04

eering

0.97

naire

0.82

jud

0.80

aries

0.78

 debtor

0.76

al

0.74

Activations Density 0.029%

expressions related to making judgments or judgments themselves

occurrences of the word "judgment" and its variants

No Comments

No Known Activations