INDEX

Explanations

phrases related to staying close or remaining committed to a course of action

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

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

 descending

-0.64

 dash

-0.62

geries

-0.62

Zi

-0.60

 brim

-0.60

 Macintosh

-0.60

 Ferguson

-0.59

 converter

-0.59

 soph

-0.57

 negotiators

-0.57

POSITIVE LOGITS

eper

0.84

leep

0.83

cation

0.83

lishes

0.82

cool

0.73

owship

0.72

bnb

0.68

til

0.68

iral

0.67

 indefinitely

0.67

Activations Density 0.039%

phrases related to staying close or remaining committed to a course of action

No Comments

No Known Activations