INDEX

Explanations

proper nouns and identifiers related to characters or entities in various 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_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

 millenn

-0.82

 textbooks

-0.78

 UNESCO

-0.77

 Ministers

-0.76

 transact

-0.74

 counseling

-0.72

 pathology

-0.72

 credential

-0.72

 foundational

-0.71

 closures

-0.71

POSITIVE LOGITS

Gaming

1.15

bear

1.15

1.11

Magikarp

1.08

guy

1.07

1.06

pie

1.04

ragon

1.00

0.98

 Profile

0.98

Activations Density 0.121%

proper nouns and identifiers related to characters or entities in various contexts

No Comments

No Known Activations